| 张猛,刘秀成*,王建国,何存富.列车车轮钢表面硬度的微磁定量检测方法[J].实验力学,2023,38(4):425~434 |
| 列车车轮钢表面硬度的微磁定量检测方法 |
| Micromagnetic and quantitative evaluation method for surface hardness of train wheel steel |
| 投稿时间：2022-09-06 修订日期：2022-12-31 |
| DOI：10.7520/1001-4888-22-219 |
| 中文关键词: 车轮钢 微磁检测 表面硬度 BP神经网络 |
| 英文关键词:wheel steel micromagnetic testing surface hardness BP neural network |
| 基金项目:国家自然科学基金项目(11527801, 11872081, 12122201) |
|
| 摘要点击次数: 2683 |
| 全文下载次数: 48 |
| 中文摘要: |
| 多功能微磁检测技术在铁磁性材料力学性能的无损表征中具有良好的应用前景。本文研究该技术对CL65、ER7车轮钢表面硬度的无损检测能力。首先，利用变异系数分析方法评价了多功能微磁仪器对车轮钢微磁参量的重复检测性能；其次，开展了标定实验，基于斯皮尔曼秩相关系数分析了微磁参量对车轮钢表面硬度的单调表征能力；最后，通过融合多项磁参量，建立了基于BP神经网络的车轮钢表面硬度的微磁定量预测模型。外部校验结果显示，BP神经网络模型对CL65、ER7车轮钢表面硬度的预测平均误差分别约为0.59%和1.22%。 |
| 英文摘要: |
| Multifunctional micromagnetic detection technology has a good application prospect in the non-destructive characterization of mechanical properties of ferromagnetic materials. In this paper, the non-destructive testing ability of this technology on the surface hardness of CL65 and ER7 wheel steel is studied. Firstly, the coefficient of variation analysis method was used to evaluate the repeated detection performance of multifunctional micromagnetic instruments on the micromagnetic parameters of wheel steel. Secondly, calibration experiments were carried out, and the monotonic characterization ability of micromagnetic parameters on the surface hardness of wheel steel was analyzed based on the Spearman rank correlation coefficient. Finally, by fusing a number of magnetic parameters, a micromagnetic quantitative prediction model for wheel steel surface hardness based on BP neural network is established. The external verification results show that the average error of the BP neural network model for the surface hardness of CL65 and ER7 wheel steel is about 0.59% and 1.22%, respectively. |
| 查看全文 下载PDF阅读器 |
| 关闭 |
